Altera launches AGILEX 3 FPGAS for the intelligent edge

Rate this post

Al thera launched its most recent family on Agilex FPGA 3 The chips to help developers extend the boundaries of the intelligent edge.

The Agilex 3 FPGA chips are already in production, said Sandra Rivera, CEO of Altera, at a briefing at the press. These devices are important for increasing the intelligence of the edge and allowing the spread of AI.

ALTERA Gate Arrame (FPGA) programmable chips are programmable chips that have found a home in a wide variety of end applications, including robotics, factory automation systems and medical equipment. ALTERA is aimed at its latest Agilex FPGAS, as well as its Quartus Prime Pro Software and FPGA AI Suite on the Intelligent Edge market.

This is a big start for Altera as it is the first as a newly independent company. Altera was originally a chip manufacturing company that was acquired by Intel in 2015 for $ 16.7 billion and merged into the company data block under the Programmable Solutions Group (PSG). But Intel has faced a number of production challenges and changes in leadership.

After nearly a year of preparation, Altera officially turned from Intel on January 10, with longtime CEO of Intel Rivera as CEO.

Sandra Rivera is CEO of Altera.

Agilex 3 FPGAS can supply 1.9 times higher tissue performance than the previous generation with 38% lower power, Rivera said, in a briefing in the press.

“Since last year, we were really focused on our leadership road card and excellent achievements, and we really drive a full -stroke portfolio for all the segments and clients we maintain and that we serve,” Rivera said. “And this really starts with this waterfall strategy that we have for the entire AGILEX road map, which we have introduced on the market. We are now introducing the latest component of this regarding the AGILEX road card. This is more limited by the cost of the traffic card power. “

A large family

The Agilex 3 FPGA has dealt with the low class of energy systems.
The Agilex 3 FPGA has dealt with the low class of energy systems.

At the highest end, Altera offers its Agilex 7 and Agilex 9 CIHP for high -density of calculations. It has the AGILEX 5 in the middle range and now the Agilex 3 ends the product line.

The recent introductions of AGILEX 5 and Agilex 3 more barriers to enter, increase market participation and simply accelerate the degree of innovation in general in the programmable logical industry, Rivera said. At the highest end, the Agilex 7 and Agilex 9 are suitable for infrastructure, network and communication applications, as well as for aerospace sectors with the built-in RF and a lot of chip.

“In the middle range of applications, we have capabilities to optimize power and performance and supply twice as well as the performance of watt as competition in this very meat, large medium range of applications and clients and segments,” Rivera said.

She said that ALTERA programmed solutions meet the strict requirements for power, performance and size of built -in and intelligent applications on the edge. These hardware solutions, along with Altera’s FPGA AI Suite, enable machine learners, software developers and FPGA designers to create custom FPGA AI platforms using standard frames such as Tensorflow and Pytorch,
and development tools like Openvino and Quartus Prime Software.

Rivera said the AGILEX 3 FPGA optimized for costs are available. FPGA’s high -performance programmable architecture, together with built -in AI tensor blocks and built -in processors, enables the business to quickly modernize its edge and built -in infrastructure by deploying personalized AI solutions that provide low latency, energy efficiency and agility needed for system long.

In the AGILEX 3 FPGA robot management systems, they provide real -time control of multi -body robot weapons by merging machine learning capabilities into many sensory pipelines. And in intelligent factory cameras, the Agilex 3 FPGA improves the detection of defects by using fine -grained parallel treatments and CNN, trained to recognize objects for analysis of huge amounts of data.

I asked Rivera if it was concerned that TSMC was affected by the possible tariff wars if the trade war flows into Taiwan, as Donald Trump said. She noted that since Altera uses both TSMC and Intel Foundry (Intel’s production contract), it is protected from such concerns. Intel has numerous chip factories in the US, and TSMC also makes chips in the US now. And in particular, projects related to Altera’s defense also require domestic production and this will not be affected.

How are FPGAS used in the field

Agilex FPGA can be used in intelligent factories.

For developers and customers, this means that Altera offers more intelligent calculation of industrial and built -in markets. This means that the intelligence of AI can spread far into everyday end devices where more profitable and network calculations can be carried out.

Devices can enhance applications, including industrial, car and defense users. This can help the drones be located with more integrated intelligence as an example. It also allows for seamless integration of sensors and triggering mechanisms in robotics in smart factories. These robots require accurate engine control algorithms to effectively function the factory floors.

“FPGA will be widely used in these environments as they can fit multiple operations in parallel and effectively deal with this continuous flow of data streams,” Rivera said. “So using the FPGA that in many cases it is already there and designed in the system, we see our built -in developers combine machine learning capabilities in their maintenance pipeline or their touch pipeline to include AI algorithms in the edge robotics.”

FPGA are also used in endoscopy and MRI machines in the medical fields. And they can bring more computing power and efficiency in the processing of data from lidar sensors in self-driving cars. And since the chips are reprogrammed in the field, they can be updated by air to adjust for the development of standards, algorithms or capabilities. This makes the design future proof, Rivera said.

Ready

Increasing the intelligence of the edge is key to the spread of AI.
Increasing the intelligence of the edge is key to the spread of AI.

Customers can now place orders for production quality Agilex 3 devices, development kits, selected partner boards and system modules.

The first wave of AGILEX 5 FPGAS electronic series devices is already fully qualified and is launched for high -volume production. Compared to the AGILEX 5 D-Series FPGAS, the Agilex 5 E-Series FPGA are optimized for more power-sensitive applications that require high-efficiency with smaller shape factors and logical density. Agilex 5 E-Series FPGA, with AI-Infused Tissue, carry high levels of integration and improved computing capabilities for intelligent EDGE applications, including video, industrial, robotic and medical
Systems.

In constant efforts to improve the cost -optimized product portfolio, Altera expands the Max 10 FPGA family with new package options. Max 10 10M40 and 10M50 product lines are available in Pitch BGA variable packages. This new package option significantly increases the value of these highly integrated devices by reducing the formal factor while maintaining a large number of IO, leading to a lower total property price for consumers. Customers can now place orders for engineering samples from Max 10 FPGA in the VPBGA-610 package, with the production silicon available through Q3 2025.

“With today’s reports, we continue to expand our leadership programmable portfolio, offering an even wider range of end-to-end solutions, built in decades of expertise and a strong ecosystem of partners,” Rivera said. “With our latest FPGA development products and tools, we provide built-in developers with a hassle-free approach to providing high-performance and high-quality intelligent solutions for the edge of the AI ​​era.”

ALTERA presents its latest FPGA innovations and development tools at Embedded World 2025 in Nurenburg, Germany.


 
Report

Similar Posts

Leave a Reply

Your email address will not be published. Required fields are marked *